Skip to main content
. 2007 Mar 22;80(5):948–956. doi: 10.1086/513477

Table 2. .

PCoA Significance Tests[Note]

All SNPs(n=9,111)
SNPs >50 kb Apart(n=6,349)
SNPs >100 kb Apart(n=5,555)
Test Adjusteda r2 (P) SKT Pb Adjusteda r2 (P) SKT Pb Adjusteda r2 (P) SKT Pb
PC1 .90 (<.001) <.0001 .89 (<.001) <.0001 .90 (<.001) <.0001
PC2 .78 (<.001) <.0001 .74 (<.001) <.0001 .72 (<.001) <.0001
PC3 .43 (<.001) <.0001 .50 (<.001) <.0001 .35 (<.001) <.0001
PC4 .54 (<.001) <.0001 .30 (<.001) <.0001 .19 (<.01) <.01
PC5 <.1 (NS) NS <.1 (NS) NS .13 (<.05) <.001
PC6 <.1 (NS) NS <.1 (NS) NS .18 (<.01) <.001
PC7 <.1 (NS) NS .17 (<.01) NS .18 (<.01) <.01
PC8 <.1 (NS) NS <.1 (NS) NS <.1 (NS) NS
PC9 <.1 (NS) <.01 <.1 (NS) NS <.1 (NS) NS
PC10 .13 (<.05) NS <.1 (NS) NS <.1 (NS) NS

Note.— Significance tests using the ANOVA and SKT. PCoA was conducted separately for each SNP set. The French singleton and the German outlier were excluded. Percentages of the variance explained by each PC are generally the same in all three cases (fig. 4A). NS = not significant at the .05 level.

a

ANOVA correlation coefficient (adjusted r2) in bold.

b

Combined P value calculated for SKT, as described in the text.